By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals/year while coaching agents nationwide.

Free AI is good enough for most of what a real estate agent does daily — drafting listing copy, social captions, follow-up emails, and market summaries. Free tiers break on three things: client data privacy, rate limits during a listing launch, and pushing output into your CRM. This guide covers exactly where that line sits.

Key Takeaways

• Free AI handles roughly 90% of an agent’s daily writing work at a quality level clients never notice.
• The upgrade decision is a privacy and integration decision, not a quality decision.
• Paying $240 a year against $9,530 in median annual business expenses is not the reason your AI isn’t working.
• Free tiers are chat windows — they don’t push into your CRM, run on a schedule, or touch an API.
• If your AI output dies in the chat window, upgrading the plan changes nothing.

What is “free AI” for real estate agents?

Free AI means the no-cost tier of a general-purpose assistant — ChatGPT, Claude, Gemini, Copilot — plus the free versions of adjacent tools like Canva and CapCut. You get access to a capable model through a chat window or app, usually with limits on how many messages you send, how large a file you upload, and which advanced features you can reach. Paid tiers lift those limits and add integrations.

ChatGPT dominates agent usage at 58%, followed by Gemini at 20% and Copilot at 15%, according to NAR’s 2025 Technology Survey. Most agents are already using one of these. The question isn’t which one — it’s whether the free version is holding you back.

Why this matters for real estate agents

Here’s the thing nobody wants to tell you: the subscription isn’t what’s broken. AI adoption reached 68% of agents in 2025, yet only 17% reported a significant positive impact on their business and 46% saw no noticeable difference at all, per HousingWire’s coverage of NAR’s 2025 Technology Survey. That 51-point gap between “using it” and “getting something out of it” doesn’t close with a credit card.

The cost argument doesn’t survive the math either. Median total business expenses for the typical Realtor climbed to $9,530 in 2025, up from $8,010 the year before, according to NAR’s 2026 Member Profile. A $20-a-month AI subscription runs $240 a year. That’s 2.5% of what you’re already spending. Meanwhile 34% of agents report spending $50 to $250 a month on tech tools already.

If $240 is the deciding factor in your business, the tool isn’t the problem. The absence of a system is.

Where free AI is genuinely good enough

Can free AI write listing descriptions?

Yes, and the quality difference is not what agents think it is. Listing copy, MLS remarks, and property highlight bullets are short-form writing tasks against structured input — exactly what every current model handles well at every tier. The output still needs your local knowledge layered on and every factual claim verified against MLS and tax records. That verification step is your job whether you paid or not.

Can free AI handle social captions and repurposing?

Yes. Turning one listing into a week of captions, hooks, and email subject lines sits comfortably inside free limits, and free tiers of editing tools cover the production side — both CapCut and Descript include AI auto-captioning without a subscription, which is the backbone of the batch video system I teach. One filming session, one editing session, a week of content. No paid plan required.

Can free AI do market research and prep?

Yes, with a caveat. Summarizing a neighborhood, building a listing presentation outline, or rehearsing an objection-handling script are all strong free-tier uses. The caveat is currency — models don’t know your market’s last 30 days, and pulling live data is where paid features start earning their keep. Prep with AI. Price with your MLS.

Where free AI actually breaks

None of these are about writing quality. That’s the part agents get wrong.

Client data and privacy

This is the only reason to upgrade I’d defend in front of a room. The moment you paste a seller’s financials, an inspection report, a lead list, or anything with client names into a chat window, you’ve stopped asking “is this good writing” and started asking a brokerage risk question. Free consumer tiers and paid or business tiers can differ on whether your inputs are retained or used to improve the model — read the terms for the specific tool and tier you’re on, and check your brokerage’s policy before you paste anything.

Concern about this is widespread, not fringe. Among agents surveyed by RPR, 63% cited accuracy of outputs as their top AI concern, 49% flagged compliance or legal issues, and 28% named fair housing, as reported by HousingWire in February 2026.

Rate limits at the worst possible moment

You don’t hit the cap on Sunday night while you’re experimenting. You hit it Thursday afternoon during a listing launch, with the photographer’s files in one tab and a seller waiting on the marketing plan. Free tiers throttle. That’s a reliability problem, not a capability problem — but reliability is what a system runs on.

No destination

Free tiers are chat windows. They don’t write into Follow Up Boss, don’t fire on a trigger, don’t touch an API, and don’t run while you’re at a closing. Everything worth automating around your listings — saved-search alerts, CRM sync, comp pulls, marketing generated from listing data — lives on the integration side. I’ve mapped what’s actually automatable around the MLS and what isn’t, and almost none of it happens inside a free chat interface.

“Free AI didn’t stall your business. A chat window with no destination did. If the output doesn’t land in your CRM or your marketing template, you didn’t build a system — you built a hobby.”
— Emily Terrell, Tom Ferry Coach

How I use this in my own business

I close 70+ transactions a year in San Antonio on roughly five hours a week of active management, and I run a mix of free and paid AI on purpose. My listing marketing suite — MLS remarks, three social captions, an email to my sphere, and a video script — gets built in one prompt, and I’ve built that suite in a free tier more than once while sitting in the car line.

What’s on a paid plan is narrow and deliberate: anything touching client data, and anything that has to run on a schedule into my CRM. When I took a Stone Oak listing to market this spring, the copy came out of a chat window in about two minutes. What made it work wasn’t the model — it was that the output had somewhere to go. My transaction coordinator has a template waiting for it. That’s why the work survives a week where I’m on a stage instead of at my desk.

The agents I coach who upgraded first and built the workflow second are still stuck. The order matters. That’s the same principle behind what real AI training has to cover — build against your own material, or nothing changes in 30 days.

What about compliance?

Your tier doesn’t change your obligations. AI-generated marketing is still advertising, which means Texas license holders still need the name of the license holder or team placing the ad plus the broker’s name in a readily noticeable location, sized per TREC’s advertising rules under Rule 535.155. A caption drafted by a model and posted without broker identification is a violation the same as one you typed yourself.

Fair Housing applies identically. Models will happily produce language describing the buyer instead of the property — “perfect for a young family,” “great for empty nesters,” “walkable to churches” — and that copy is your liability, not the tool’s. Read every AI-generated listing description for protected-class references before it goes anywhere.

This is general information, not legal advice. Confirm your specific advertising and data-handling practices with your broker and an attorney licensed in your state.

Common mistakes

Upgrading to fix an output problem. If the copy sounds generic, that’s a prompt and input problem. Better inputs beat a better plan every time.

Pasting client data into a consumer free tier. The fastest way to turn a time-saver into a brokerage conversation you don’t want to have.

Tool-shopping instead of workflow-building. Trialing six tools in a quarter produces zero workflows. Pick one, learn it well, and wire it into something.

Publishing AI copy unverified. Every number, every school rating, every square footage gets checked against MLS and tax records before it is published.

Treating the chat window as the finish line. If the output doesn’t move into a template, a CRM, or a scheduled post, you’re doing arts and crafts. This is also why most agents stay invisible in AI search — the fix is structural, not volume.

Frequently Asked Questions

Is free ChatGPT good enough for real estate agents?

For drafting listing descriptions, social captions, follow-up emails, and scripts, yes. Free-tier output quality on short-form real estate writing is close enough that clients won’t notice a difference. Where free ChatGPT falls short is rate limits during high-volume weeks, handling of sensitive client data, and any connection into your CRM or automation stack.

What can’t free AI tools do for real estate agents?

Free AI tools generally can’t connect to your CRM, run on a schedule, process large document uploads reliably, or offer the data-handling terms most brokerages want for client information. They also throttle usage. Everything on that list is an integration or reliability limit — not a writing-quality limit, which is where most agents wrongly assume the gap lives.

Should I put client information into a free AI tool?

Check the terms of your specific tool and tier, plus your brokerage’s policy, before you do. Consumer free tiers and business or paid tiers can differ on whether inputs are retained or used for model training. The practical rule: strip names, addresses, financials, and contact details out of anything you paste, or use a tier your broker has approved.

How much should a real estate agent spend on AI tools?

Start at zero and let the workflow tell you when to pay. Most agents can run their entire content operation on free tiers for months. For context, 34% of Realtors spend $50 to $250 monthly on all tech combined, per NAR’s 2025 Technology Survey. Upgrade when you hit a specific wall — data handling, rate limits, or integration — not on principle.

Does paid AI write better listing descriptions than free AI?

Marginally, and not in ways buyers notice. The larger quality difference comes from your input: a detailed property brief with specific features, neighborhood context, and a target buyer produces better copy on a free tier than a vague one-line prompt produces on a paid tier. Fix the prompt before you fix the plan.

When should a real estate agent upgrade to a paid AI plan?

Upgrade when one of three things happens: you need to process client data under terms your broker approves, you’re hitting usage limits during listing launches, or you’re ready to connect AI output into your CRM or automation stack. All three are workflow triggers. If none apply, the paid plan buys you nothing you’ll use.

Is free AI safe to use for Fair Housing compliance?

No AI tier is Fair Housing compliant on its own — compliance is your review, not the tool’s setting. Models routinely generate copy describing the ideal buyer rather than the property, which is exactly the language Fair Housing prohibits. Read every AI-drafted listing description for protected-class references before publishing, regardless of what you’re paying.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals a year while coaching agents nationwide.

When a client brings you a Zestimate, don’t argue with it — open Zillow’s own accuracy page in front of them. Zillow publishes a much higher median error rate for off-market homes than on-market ones, and your seller is reading the off-market number. Here’s the Zestimate objection script, the data behind it, and the system that prevents it.

Key Takeaways

• The Zestimate your seller checked before your appointment is the off-market version, which Zillow’s own data shows is several times less accurate than the on-market version.
• Arguing with the number makes you look defensive and makes the seller protective of it. Showing them Zillow’s published accuracy data does the work for you.
• Zillow itself states the Zestimate is not an appraisal and points consumers toward a CMA from an agent.
• The on-market accuracy number looks strong partly because listing data, including list price, feeds the model once a home goes live.
• If the objection is happening at the listing table, your pre-appointment system failed. Move the conversation into your pre-listing package.

What is a Zestimate, and why does it show up in listing appointments?

A Zestimate is Zillow’s automated estimate of a home’s market value, generated by a neural network model that pulls county and tax assessor records, direct MLS feeds, home characteristics, prior sales, and market trend data. Zillow publishes Zestimates for more than 110 million homes and refreshes them multiple times per week. (Zillow, What is a Zestimate?)

That scale is why it shows up in your listing appointments. Your seller has been watching their Zestimate move for years before they ever considered calling an agent. By the time they sit down with you, that number isn’t data to them. It’s an expectation.

Zillow is direct about the limits. The company states plainly that the Zestimate is not an appraisal and cannot be used in place of one, and it encourages buyers, sellers, and homeowners to supplement the estimate with other research — including requesting a comparative market analysis from a real estate agent. (Zillow, What is a Zestimate?)

Read that again. The company that produced the number your seller is quoting points them toward your CMA as the next step. That’s not an argument you have to construct. It’s already published.

Why the Zestimate objection matters for real estate agents

Because the moment goes badly for most agents, and the seller notices.

An agent presents a carefully built CMA at $485,000. The seller pulls out their phone: “Zillow says $512,000.” Then either silence, or the agent starts walking back their own analysis to stay in the room. That’s a $27,000 credibility gap opened by a number nobody in the room understands.

Here’s the thing nobody wants to tell you: the seller isn’t testing the price. They’re testing whether you know more than the algorithm. If you flinch, you’ve answered.

The market data says that the test is winnable. Ninety-one percent of home sellers used a real estate agent or broker to sell their home, and for-sale-by-owner transactions dropped to a historic low of just 5% of sellers. (NAR, Top 10 Takeaways from the 2025 Profile of Home Buyers and Sellers, November 2025)

Automated valuations have been public for nearly twenty years and sellers are using agents at record rates. The AVM didn’t replace you. It changed what the first ten minutes of your appointment are about.

The number your seller is reading is the least accurate one

This is the part most agents miss, and it’s the entire conversation.

Zillow publishes two different accuracy figures, because the Zestimate behaves differently depending on whether a home is listed. The nationwide median error rate sits under 2% for on-market homes and around 7% for off-market homes — a gap of roughly four times. Zillow breaks both down by region on its own accuracy tables. (Zillow, What is a Zestimate?)

Why the gap? Zillow explains it directly: for on-market homes, the model takes in listing information including the list price and recent market activity, and that information simply isn’t available while a home is off-market.

So the on-market number gets a boost from data an agent generated. And the off-market number — the one with fewer inputs and the wider range — is the one every unlisted homeowner in your market is looking at right now.

Run the math for your seller, out loud, in their house. On a $650,000 Stone Oak home, a 7% median error is roughly $45,500. And “median” means half of all estimates miss by more than that. That’s not a rounding error. That’s a two-bedroom condo.

Then point at the Estimated Sale Range on their own listing page. Zillow publishes it right there, and a wider range is the model signaling it’s less certain about that home. Most sellers have never once looked at it.

“The Zestimate isn’t your competition. It’s your opening slide. The agent who explains the number better than the algorithm can win the pricing conversation before they ever present a comp.” — Emily Terrell, Top Coach and Speaker at Tom Ferry International

The four-move Zestimate objection script

No memorized paragraph. Four moves, in order, that work in any market.

Move 1: Ask to see it

“Good — pull it up. I want to look at it with you.”

You’ve just done the one thing the seller didn’t expect. Every agent they’ve talked to either dodged the number or attacked it. You asked for it. Curiosity beats defense, and it costs you nothing.

Move 2: Open Zillow’s own accuracy page

Don’t tell them the Zestimate is wrong. Show them what Zillow says about it. Keep Zillow’s Zestimate page bookmarked on your phone, and walk them through the on-market versus off-market error rates and the line where Zillow says it isn’t an appraisal.

You’re not attacking Zillow. You’re demonstrating that you understand their tool better than they do. That reframe is the whole play, and it lands in under ninety seconds.

Move 3: Name what the model can’t see

Get specific to their property. Generic beats nothing, but specific closes.

“It knows your square footage and your bed-bath count. It doesn’t know your roof is three years old, that your kitchen was redone without a permit so it never hit county records, that you went back to Greenbelt instead of the retention pond two streets over, or that the comp it’s leaning on closed at a discount because of a divorce.”

Four concrete things the algorithm missed. Now the seller is doing the math with you instead of against you.

Move 4: Put your CMA beside it, not against it

Never position your number as a correction. Position it as the finished version.

“Both of these are starting points. Only one of them has been inside your house.”

Then present. Your analysis isn’t replacing the AVM — it’s completing a picture the model can’t finish on its own. Sellers accept completion. They resist correction.

What about Redfin, Realtor.com, and bank estimators?

Same architecture, same ceiling. Redfin, Realtor.com, and the bank-branded home value tools all run machine learning models over public records, tax assessments, and recent sales — the same inputs, with the same blind spot about what’s actually inside the house.

The tactical version: pull three AVMs for the same property in front of the client. When one says $512,000, another says $486,000, and a third says $530,000, you no longer have to argue that any single one is wrong. A spread that wide on the same house makes the point for you.

That demonstration takes ninety seconds and it ends the objection permanently, because the seller has now watched the tools disagree with each other.

How I use this in my own business

I sat down with a Stone Oak seller last fall whose Zestimate was running about $40,000 above where the comps supported. They’d been watching that number for two years. It was the first thing they said when I walked in.

I didn’t touch the price for the first ten minutes. I pulled up Zillow’s accuracy page on my phone, showed them the off-market error rate, and applied the percentage to their own address. Then I pulled the Estimated Sale Range on their listing page — it was more than $80,000 wide. They’d never noticed it.

We listed inside my range. It went under contract in the first weekend.

The point isn’t that I talked them down. It’s that I never had to. Zillow’s own disclosure did the work, and I stayed the person in the room who understood the data.

The system that prevents the objection entirely

Here’s the uncomfortable version: if the Zestimate is coming up during your listing appointment, your pre-appointment system already failed.

By the time a seller says the number out loud in their living room, they’ve anchored to it emotionally, and you’re now the person taking something away from them. The script above rescues that moment. A system means you never need it.

The pre-listing insert

Build a one-page PDF titled something like “What your online home estimate can and can’t see.” Three sections: how AVMs generate a number, the on-market versus off-market accuracy gap with Zillow’s published figures, and a short list of what no model can evaluate without walking the property.

Send it with your pre-listing package, forty-eight hours before the appointment. The seller reads it alone, with no agent in the room to argue with, and does the recalibration privately. You walk in and the objection is already handled.

That’s the difference between a script and a system. The script is you performing. The system is you not having to.

The automation

Put the comp pull and the insert on a trigger, so a booked listing appointment fires both without anyone remembering to do it. This is the same principle behind automating the workflow around your MLS — automate the gathering and the formatting, keep your hands on the judgment.

Scalable and repeatable. Build it once, and it runs on every appointment for the next five years.

Common mistakes

Attacking Zillow. The second you criticize the platform, the seller gets protective of the number. Let Zillow’s own published data disqualify the estimate for you.

Quoting a number you didn’t verify. Zillow updates its accuracy figures. Pull the live page in the appointment rather than reciting a statistic you memorized eight months ago. Being caught with a stale number costs more than not having one.

Treating the on-market rate as the relevant one. Agents who quote the sub-2% figure are quoting the number that doesn’t apply to their seller’s situation and accidentally arguing against themselves.

Making it about accuracy instead of completeness. “The Zestimate is wrong” invites debate. “The Zestimate is incomplete” invites you to complete it.

Handling it live every single time. If you’re running the same objection at every appointment, you have a content problem, not a conversation problem. Move it upstream.

Using the word “value” about your own analysis. In Texas, that’s a licensing issue, not a style preference. See below.

What Texas agents need to know about calling it a “value”

This is general information, not legal advice. Confirm specifics with your broker or attorney.

Texas draws a hard line here. Under TREC Rule 535.17, a real estate license holder may not perform an appraisal of, or provide an opinion of value for, real property unless that license holder is licensed or certified as an appraiser under Texas Occupations Code, Chapter 1103. (TREC Rules)

The same rule requires that when you provide a broker price opinion, comparative market analysis, or estimated worth or sale price, you also give the client a written statement — reproduced verbatim, in at least 12-point font: “This represents an estimated sale price for this property. It is not the same as the opinion of value in an appraisal developed by a licensed appraiser under the Uniform Standards of Professional Appraisal Practice.”

Two details that catch agents:

The rule covers “estimated worth or sale price,” not just documents labeled CMA. That’s broader than most agents assume, so if you’re handing a client any price estimate as part of your presentation, check with your broker about whether the disclosure follows it.

And a sales agent must submit the CMA in the sponsoring broker’s name — the broker is responsible for it.

Practical translation: say “estimated sale price,” not “value,” when you’re describing your own analysis. Get the disclaimer on the document. Most CMA platforms insert it automatically, but check yours rather than assuming.

Frequently Asked Questions

How accurate is a Zestimate really?

It depends entirely on whether the home is listed. Zillow publishes a nationwide median error rate under 2% for on-market homes and around 7% for off-market homes. Median means half of all estimates miss by more than that figure. Accuracy also varies significantly by metro area, and Zillow publishes regional tables showing the spread.

What should I say when a seller quotes their Zestimate?

Ask to see it. Then open Zillow’s own accuracy page and walk them through the off-market error rate applied to their specific address. Name three or four things about their property the model can’t see. Then present your CMA as the completed picture rather than a correction. Curiosity beats defensiveness every time.

Why is my client’s Zestimate different from my CMA?

Because the model has never been inside the house. Zestimates run on public records, tax assessments, prior sales, and home characteristics. They can’t evaluate condition, finish quality, unpermitted renovations, micro-location advantages, or the circumstances behind a discounted comp. Your CMA incorporates all of that, which is why the numbers diverge.

Is a Zestimate an appraisal?

No. Zillow states directly that the Zestimate is not an appraisal and cannot be used in place of one, and that it can’t be used to get a loan because lenders rely on professional appraisals. It’s a computer-generated estimate produced with no physical inspection of the property.

Why does the Zestimate change when a home goes on the market?

Once a home lists, the model gains access to listing data — including list price and recent market activity — that isn’t available while the property is off-market. Zillow confirms this on its own Zestimate page. That’s a meaningful part of why the on-market accuracy figures look stronger than the off-market ones.

Can an agent get a Zestimate changed?

Not directly. The Zestimate is generated by an automated software process and can’t be manually altered for a specific property. What you can do is help your client update their home facts on Zillow — square footage, bed and bath count, completed renovations — since unreported additions and remodels aren’t reflected in the estimate. Reporting updates to the local tax assessor helps too.

Should I put my CMA number in writing before the appointment?

Send the education, not the number. A pre-listing insert explaining how AVMs work and where they break down recalibrates the seller before you arrive. Your actual pricing recommendation belongs in the appointment, where you can present comps, answer objections, and read the room.

What’s the required disclaimer on a CMA in Texas?

TREC Rule 535.17 requires specific language, reproduced verbatim in at least 12-point font, stating that the document represents an estimated sale price and is not the same as the opinion of value in an appraisal developed by a licensed appraiser under USPAP. The rule applies to broker price opinions, CMAs, and estimated worth or sale price. Confirm current requirements with your broker.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. If you’re deciding what kind of speaker your event actually needs, start with the difference between a keynote speaker and a motivational speaker. Recent stages include NAHREP and eXp Con. More about Emily.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals/year while coaching agents nationwide.

Real estate AI tools cost most agents between $50 and $500 a month, and NAR’s 2025 Technology Survey found 46% of agents saw no noticeable impact from AI. The problem is stack sprawl, not tool quality. This post gives you a four-question audit that decides what to keep and what to cancel this week.

Key Takeaways

• Most agents can’t name the task each AI subscription removed — that’s the audit failure, not the tool’s failure.
• Two-thirds of agents say their brokerage already provides all the tech they need, yet 44% still spend $251 or more per month.
• Your CRM may already include AI features you’re paying a second vendor to duplicate.
• Four questions decide every renewal: what task, what duplicate, what usage, what breaks.
• Cancel first, then rebuild. Adding tools to a broken system just makes the mess more expensive.

What is an AI stack audit?

An AI stack audit is a line-by-line review of every AI and tech subscription you pay for, scored against the specific recurring task each one removes from your week. It’s not a comparison of features. It’s a comparison of cost against hours returned. Anything that can’t name its task gets cancelled.

Why this matters for real estate agents

The money is already going out the door. According to NAR’s 2026 Member Profile (June 2026), median total business expenses rose to $9,530 in 2025, up from $8,010 the year before. That figure comes straight from NAR’s release on the 2026 Member Profile, which also reports median gross income of $59,200 and a typical nine transaction sides per individual agent. Read NAR’s 2026 Member Profile release TrustRadius

Now look at what that spend is buying. According to NAR’s 2025 Technology Survey (September 2025), 34% of agents spend between $50 and $250 per month on tech tools, 20% spend $251 to $500, and 24% spend over $500 — while 46% reported that AI had no noticeable impact on their business and only 17% reported a significantly positive impact. See NAR’s 2025 Technology Survey findings HousingWire

Here’s the thing nobody wants to tell you: nearly a quarter of agents are spending over $500 a month on tools, and almost half of the profession reports getting nothing measurable back. That’s not a technology problem. That’s a purchasing problem.

There’s a third number that makes it worse. Two out of three agents either agree (38%) or strongly agree (29%) that their brokerage already provides all the tech tools they need. Most agents are buying a second copy of something they already have. Full 2025 Technology Survey report HousingWire

The four-question AI tool audit

Pull up your bank statement and your credit card statement. List every recurring software charge. Then run each line through these four questions in order.

Question 1: What named task does this remove?

Write the task in one sentence, in the present tense, using a verb. “Drafting my listing description from MLS data” is a task. “Helps with marketing” is not. If you can’t finish the sentence in ten seconds, the tool doesn’t have a job — it has a subscription.

Question 2: Are you already paying for this somewhere else?

This is where most of the waste lives. Follow Up Boss, for example, publishes AI features as part of every plan tier — smart summaries, smart messages, suggested tasks, and predictive lead prioritization — with no separate AI charge. Check what’s included in Follow Up Boss AI If you’re paying for a standalone AI writing tool to draft follow-up messages while your CRM already does it, cancel one. Same logic applies to your brokerage’s included tech and your MLS’s included tools.

Question 3: Did you open it in the last 30 days?

Not “did you mean to.” Did you open it? Log into each tool and check the last-activity date. Anything untouched for 30 days is a cancellation, not a maybe. You can always resubscribe — most of these have no contract and month-to-month billing. Follow Up Boss publishes its tiers and add-on costs openly, which makes the math easy to run before you renew.

Question 4: What breaks if you cancel it tomorrow?

Name the consequence. “My transaction coordinator loses the checklist” is a real consequence. “I’d feel behind” is not. If nothing breaks, nothing is working.

What top-producing agents actually keep

Strip the stack down and the agents producing consistently keep four layers, not fourteen.

The thinking layer

One general AI model for drafting, summarizing, and structuring. NAR’s 2025 Technology Survey found ChatGPT was the most-used AI tool among agents at 58%, followed by Gemini at 20% and Copilot at 15%. You need one. Not three. Pick the one whose output needs the least editing in your voice and stop tool-shopping. HousingWire

The follow-up layer

Your CRM, with its AI features, is actually turned on. This is the layer that touches revenue, and it’s the one most agents underuse while overbuying around it. If you’re evaluating this layer, start with the AI-CRM stack breakdown rather than a feature comparison chart.

The video layer

One editor with AI captioning. That’s it. The gear and the editing workflow matter more than the software tier, which is why the batch editing system beats upgrading your plan.

The design layer

One template tool holding your brand assets, so you’re not rebuilding a listing graphic from scratch every time. Everything past this — the fifth AI headshot generator, the AI social scheduler that duplicates your CRM’s automation — is stack sprawl.

“A tool earns its renewal when you can name the recurring task it removed and the hour it gave back. If you can’t name the task, you’re not buying software. You’re buying the feeling of being current.”
— Emily Terrell, Top Coach and Speaker at Tom Ferry International

Common mistakes

Buying a tool before fixing the system underneath it. A tool bolted onto a broken follow-up process automates the broken process faster. Fix the sequence first.

Paying annually for something you haven’t used monthly. Annual billing discounts are real, but they’re only a discount if you use the tool. Otherwise you’ve prepaid twelve months of nothing.

Confusing “everyone’s using it” with “it works for me.” Adoption stats are not ROI. Nearly half of agents report no noticeable impact from AI — a lot of those agents are using the same tools everyone recommends.

Cancelling the CRM and keeping the accessories. The CRM is the layer with revenue attached. Cancel outward from the edges, never inward from the center.

Skipping the brokerage check. Before you buy anything, ask your broker what’s already included. Two-thirds of agents say their brokerage covers what they need, and most never check the list.

Frequently Asked Questions

How much do real estate agents spend on AI tools each month?

According to NAR’s 2025 Technology Survey, 34% of agents spend $50 to $250 per month on tech tools, 20% spend $251 to $500, and 24% spend more than $500. Those figures cover total technology spend, not AI alone. NAR does not publish AI-specific spending broken out by production level.

What AI tools do top-producing agents actually pay for?

No public dataset segments AI spending by production tier, so treat any specific list as opinion rather than data. What consistently holds is the structure: one general AI model, one CRM with AI enabled, one video editor, one design tool. NAR’s 2025 Technology Survey found ChatGPT is the most-used AI tool among agents at 58%.

Do I need a separate AI writing tool if my CRM already has AI?

Usually not. Follow Up Boss, for instance, includes AI features across all plan tiers at no additional charge, covering smart summaries, smart messages, and suggested tasks. Check your CRM’s included feature list before renewing a standalone writing tool. If both do the same job, keep the one that lives where you already work.

Is ChatGPT enough, or do I need a real estate-specific AI tool?

For drafting, summarizing, and repurposing content, a general model is usually enough — the differentiator is your prompt library, not the vendor. Real estate-specific tools earn their cost when they connect to your MLS or CRM data directly. If the specialized tool just adds a template layer on top of a general model, you’re paying twice.

How do I know if an AI tool is worth the money?

Divide the monthly cost by the number of times you used it last month. If you can’t produce that usage number, that’s your answer. Then name the recurring task it removed in one present-tense sentence. A tool that survives both tests stays. Everything else is cancellable this week.

Should my brokerage’s included tech replace what I’m paying for?

Often, yes. NAR’s 2025 Technology Survey found 38% of agents agree and 29% strongly agree that their brokerage provides all the tech tools they need. Ask your broker for the full included list before your next renewal cycle. Duplicated coverage is the single most common line item in an agent’s stack.

How often should I audit my AI stack?

Twice a year is enough. Any more and you’re managing software instead of running a business; any less and annual renewals slip through unexamined. Put both audits on the calendar as recurring appointments — one in January and one in July — and pull your actual bank and card statements rather than working from memory.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals a year while coaching agents nationwide.

Most real estate agents should budget $20 to $80 per month for AI tools — one frontier AI subscription at $20, plus video or design software only if it removes a weekly task. Your CRM is separate infrastructure, not AI spend. This guide gives you the three-tier budget and the audit that decides what stays.

Key Takeaways

• The working AI budget for most agents is $20 to $80 a month, and the floor of $20 covers the majority of what you’d actually use.
• Your CRM is infrastructure, not AI spend — counting it inflates your “AI budget” by 3x and hides what you’re really paying for.
• A tool earns renewal only when it removes one recurring task within 30 days. No named task, no renewal.
• Free tiers handle more than agents assume, especially in video editing and design.
• Median business expenses for Realtors hit $9,530 in 2025, so an $80 AI stack is roughly 10% of your annual operating cost — small enough that discipline matters more than the dollar amount.

What is an AI tool budget for real estate agents?

An AI tool budget is the recurring monthly amount you spend on software whose primary function is generating or automating work — writing, video, design, research, or communication drafting. It does not include your CRM, your MLS dues, your website, or your lead sources, even when those products ship AI features inside them. Keeping that line separate is what makes the number meaningful.

Why this matters for real estate agents

Agent operating costs are rising faster than agent income, which makes every recurring subscription a real decision. According to NAR’s 2026 Member Profile, median total business expenses climbed to $9,530 in 2025, up from $8,010 the year before, with vehicle costs the largest single category at $1,580. (National Association of REALTORS®, June 2026)

Against a median gross income of $59,200 and nine individual transaction sides, an unexamined stack of five subscriptions is not a rounding error. (National Association of REALTORS®, June 2026)

Here’s the thing nobody wants to tell you: the agents who ask “how much should I spend on AI” almost never have an underspending problem. They have five subscriptions and open two. The budget isn’t the constraint. The discipline is.

The three-tier AI budget

What does the $20 floor actually cover?

One frontier AI subscription, and nothing else. Claude Pro runs $17 per month billed annually at $200 up front, or $20 billed monthly, and ChatGPT Plus sits at the same tier. (Claude pricing, Anthropic)

That single subscription handles listing descriptions, follow-up drafts, market update copy, listing presentation prep, objection scripts, social captions, and email sequences. For an agent doing nine to fifteen sides a year, this tier is the whole answer. Start here and stay here until something specific breaks.

When do you move to the $50 to $80 working stack?

When you’re producing video weekly and the editing is the bottleneck. HeyGen’s Creator plan is $29 per month, or $24 billed annually, which adds avatar video for the repeatable content you don’t need to physically film. (HeyGen pricing)

Read the credit math before you subscribe. Creator includes 200 monthly premium credits, and Avatar IV generation consumes 20 credits per minute, so that allotment covers roughly ten minutes of premium avatar video. (HeyGen pricing)

Editing itself usually costs nothing. CapCut and Descript both have free tiers that cover most agent needs, which is exactly why the batch editing system fixes more than a paid upgrade would.

Is spending past $150 a month ever justified?

For a solo agent, rarely. Past $150 you’re either running a team, running paid ad production at volume, or you’ve stopped auditing. The ceiling isn’t a rule about money — it’s a signal to check whether every line item still maps to a task.

Team leaders are the exception. Multiple seats, shared brand assets, and a content operation change the math, and that’s a different budget conversation than the one most agents are having.

Your CRM is not AI spend

This is where agents miscount by three hundred percent. Follow Up Boss Grow is $69 per user per month, dropping to $58 billed annually, with calling as a $39 add-on ($33 annually). (Follow Up Boss pricing)

Add the dialer and a solo agent is at roughly $108 a month — but that’s CRM infrastructure you’d pay for with or without AI. Follow Up Boss includes its AI features at no additional charge; some of them just depend on data the paid Calling add-on produces. (Follow Up Boss pricing)

File it correctly and your actual AI budget is $20, not $128. That reframe alone stops most agents from believing they’re overspending on AI when they’re really just paying for a CRM. If you’re still choosing that foundation, start with how to pick an AI-ready CRM without wasting 90 days on setup.

The subscription audit that decides what stays

Open your bank statement. Filter for recurring charges. For each one, write the specific task it removed and the date it started removing it. That’s the whole audit, and it takes about eleven minutes.

“A tool earns its subscription when it removes one recurring task within thirty days. If you can’t name the task it replaced, you’re not paying for AI — you’re paying for the feeling of being current.”
— Emily Terrell, Tom Ferry Coach

Three rules make it scalable and repeatable. Cancel anything you can’t attach a task to. Never buy a second tool in the same category as one you already have. And before any new subscription, run 30 days on the free tier — if the free tier is enough, that’s your answer.

Let the platform data narrow your choices, too. NAR’s 2026 Member Profile found agents professionally use Facebook (76%), Instagram (57%), LinkedIn (55%), YouTube (31%), and TikTok (16%). (National Association of REALTORS®) You don’t need tooling for five platforms. You need it for the two you actually publish on.

How I use this in my own business

I close 70+ transactions a year in San Antonio on roughly five hours a week of active management, and my AI line item is smaller than most agents expect. One frontier AI subscription does the heavy lifting — listing copy, seller prep, market recaps, coaching materials, the weekly prompt library. Video tooling gets added seasonally when I’m producing at volume and dropped when I’m not.

Here’s the actual how. Every January I export twelve months of recurring charges, and every subscription has to justify itself against a named task. Last audit, two tools didn’t survive the question. Neither one had removed a single recurring task in a year. I’d bought both during a conference week, which is the most expensive week of the year to make software decisions.

That’s the system working. Not spend more. Fewer, better-used tools.

Common mistakes

• Counting the CRM as an AI budget. It’s infrastructure. You’d pay for it either way, and folding it in makes your AI spend look reckless when it isn’t.
• Buying at conferences. Demos are designed to be persuasive. Wait 14 days and see if you still want it.
• Subscribing before exhausting the free tier. Most agents never hit the free-tier ceiling on editing or design tools.
• Owning two tools that do the same job. One AI assistant. Not three, because a podcast recommended each of them.
• Judging tools by output quality instead of hours recovered. A tool that produces beautiful work you never use has a return of zero.
• Never auditing. A subscription you forgot about renews forever. That’s not a pricing problem — it’s a calendar problem.

Frequently Asked Questions

How much do most real estate agents spend on AI tools per month?

Most agents land between $20 and $80 a month once you exclude the CRM. A single frontier AI subscription at $20 covers writing, follow-up, and listing content. Video tools add roughly $24 to $29. Spending beyond $150 as a solo agent usually signals unaudited subscriptions rather than genuine need.

Is one AI subscription enough for a real estate agent?

For most agents, yes. One frontier assistant handles listing descriptions, client emails, market updates, social captions, scripts, and presentation prep. Adding a second general-purpose assistant duplicates capability without adding output. Add a second tool only when it does something categorically different, such as video generation or design, and only when you’re producing that content weekly.

Does my CRM count toward my AI tool budget?

No. Track it as separate infrastructure. Real estate CRMs increasingly include AI features at no extra charge, but you’d pay for the CRM regardless of whether those features existed. Counting it inflates your perceived AI spend by three to five times and makes an entirely reasonable stack look wasteful. Budget them on separate lines.

Are AI tool subscriptions tax deductible for real estate agents?

Ordinary and necessary business software is generally treated as a deductible business expense, and NAR tracks technology spend as part of members’ business expenses. This is general information, not tax advice — how it applies to your situation depends on your entity structure and use. Confirm the treatment with your CPA before you file.

How do I know if an AI tool is worth keeping?

Name the recurring task it removed and the date it started removing it. If you can’t answer both in one sentence, cancel at renewal. Output quality is the wrong test — a tool that produces excellent work you never publish returns nothing. Hours recovered is the only measure that survives an audit.

Should new agents spend anything on AI tools?

Start with free tiers for 60 days. Free plans on major AI assistants, plus free editing tools, cover more than new agents expect. Once you can name a task you do every week that a paid tier would remove, upgrade to the $20 subscription. Buying a stack before you have a workflow just adds cost to confusion.

What AI tools should a real estate team budget for?

Teams budget differently because seats multiply. Start with one shared assistant subscription per producing agent, then evaluate whether your CRM’s included AI features already cover summaries, drafting, and prioritization before adding anything on top. Most teams discover the gap is process, not software — see which real estate CRM works best with AI.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals/year while coaching agents nationwide.

The split between AI and a real estate virtual assistant is simple: give AI the repeatable, template-driven work — drafting, formatting, summarizing. Give your VA the work that needs judgment, verification, and follow-through, and keep licensed activity on your own desk. This guide gives you the three-column split, the prompts, and the compliance line you cannot cross.

Key Takeaways

• AI handles work with a predictable input and a predictable output; a VA handles work that requires a decision, a verification, or a human on the other end of the phone.
• The third column matters most: licensed activity like showings, soliciting listings, and contract review can’t go to an unlicensed assistant in Texas, no matter how well you’ve documented it.
• Most agents don’t have a delegation problem — they have a documentation problem, and AI is the fastest way to fix it.
• Hire the AI workflow first, then hire the VA to run it. Reversing that order means you’re paying a person to do work a machine already does for free.
• Review output on a schedule, not activity in real time. If you’re checking on your VA daily, you built a job for yourself, not a system.

What is the AI + VA split?

The AI + VA split is a delegation framework that sorts every task in your real estate business into three columns: work a machine can do, work a person must do, and work only a license holder can do. It replaces the vague instinct to “get help” with a written decision about where each task lives. The split is what makes a virtual assistant profitable instead of expensive.

Most agents skip this step. They hire a VA, hand over a pile of unstructured work, and then spend more time explaining tasks than they would have spent doing them. That’s not a hiring failure. That’s a design failure.

Why this matters for real estate agents

The economics are tighter than most agents admit. According to NAR’s 2026 Member Profile (June 2026), total median business expenses rose to $9,530, up from $8,010 in 2024, with vehicle costs the largest category at $1,580. The same report found that the typical individual agent closed nine transaction sides in 2025, with median gross income of $59,200. Run the math: a full-time domestic assistant consumes your entire annual expense budget several times over. The only version of this that works is a narrow, well-defined scope — and you can’t define scope without the split. Mile High Title GuyNational Association of REALTORS

Here’s the thing nobody wants to tell you. Adoption isn’t the problem anymore. An RPR survey of 225 real estate professionals published in February 2026 found that 82% said they currently use AI in their business, and 92% said they are either using AI now or planning to. But NAR’s 2025 Technology Survey reported that 46% of Realtors found a neutral or no noticeable impact on their business from AI, while 33% found a moderately positive impact. PropertyvideosAtlantaagentmagazine

Eighty-two percent are using it. Almost half report it changed nothing. That gap isn’t a tool problem — it’s a delegation problem. Agents are using AI to do faster versions of tasks that shouldn’t exist, instead of using it to redesign who does what.

“I run 70+ transactions a year in about five hours a week of active management. Not one of those hours goes to telling someone what to do. They go to review what has already been done. That’s the difference between a team and a to-do list with extra people attached.”
— Emily Terrell, Tom Ferry Coach

The Three-Column Split

Open a blank document. Three columns. Then take every recurring task in your business and put it in exactly one.

Column one: What goes to the machine?

Give AI any task where the input is predictable and the output follows a pattern. If you’d hand a human a template and say “fill this in,” you’re describing machine work.

That includes MLS listing descriptions, social captions, listing presentation copy, market update emails, blog drafts, video scripts, meeting summaries, transcribing and organizing call notes, and first-pass research on a neighborhood or a comparable. It also includes the thing most agents miss: writing the process documentation itself. A saved prompt that turns your rambling voice memo into a clean, numbered SOP is worth more than most software you’re paying for.

The test is whether a wrong answer is recoverable in under a minute. If you can glance at the draft and fix it, it’s machine work. If a wrong answer costs you a client or a commission, it isn’t.

Column two: What goes to your virtual assistant?

Give your VA any task that needs judgment, verification, or a human on the other end.

That’s transaction file management and deadline tracking, chasing title and lender for documents, scheduling, CRM hygiene, confirming that the AI-drafted listing description actually matches the property, posting and monitoring social content, vendor coordination, and inbox triage. Your VA is the verification layer on top of everything in column one. AI drafts, the VA checks, you approve.

This is also where the “managing a VA with AI” question actually gets answered. You don’t manage a person with AI. You use AI to build the documentation, checklists, and templates that make the person self-directing — then you review output on a fixed cadence instead of supervising activity in real time. The systems that keep your listing pipeline moving without you work the same way: the automation runs, the human confirms, you check the result.

Column three: What never leaves your desk?

This is the column that keeps agents out of trouble, and it’s the one almost nobody writes down.

In Texas, an unlicensed assistant — including an offshore VA — can’t perform activity that requires a real estate license, and the exposure runs both directions. According to the Texas Real Estate Commission, someone who conducts real estate brokerage activity without a license, as well as a broker or sales agent who employs that unlicensed individual, commits a Class A misdemeanor, which carries a fine up to $4,000 or up to a year in jail, or both. TREC may also take disciplinary action against a license holder who pays or associates with an unlicensed person performing licensed activity. HousingWireHousingWire

The specifics matter more than the headline. Under TREC Rule 535.4(f), an unlicensed assistant can’t make calls to determine whether someone is interested in buying, selling, or leasing property — even to schedule a follow-up appointment with a license holder. Under Rule 535.4(c), they can’t show a property, and “show” includes unlocking a door, allowing access, or hosting an open house. They can’t review a contract, though they can input data into one when specifically directed by a license holder. HousingWire + 2

What they can do is genuinely useful. TREC confirms an unlicensed assistant may arrange a showing appointment on behalf of the license holder who will show the property, input data or type contracts as specifically directed, handle bookkeeping, and order supplies or schedule maintenance. Under Rule 535.5(f), they may also confirm information previously advertised about a specific property — such as size or price — but should refer the caller to a license holder for follow-up questions the advertisement doesn’t cover. HousingWireHousingWire

Column three is your name on the line: pricing opinions, negotiation, listing appointments, contract review, and every conversation where a client is making a decision. TREC’s own guidance is that brokers should establish written guidelines and training telling agents and unlicensed personnel exactly what’s allowed. Build that document before your VA’s first day, not after their first mistake. HousingWire

This is general information, not legal advice. Licensing rules vary by state — confirm the requirements with your broker and your state commission before you delegate anything.

How do you build the split in one afternoon?

Step one: Audit two weeks backward, not forward

Open your calendar and your sent folder and list every task you actually performed in the last ten business days. Don’t plan from what you think your week looks like — plan from what it was. Most agents find 30 to 40 recurring tasks and are surprised by how many are pure formatting work.

Step two: Sort ruthlessly, then check the third column twice

Assign each task a column. Anything you hesitate on goes to column three until you can prove otherwise. Hesitation is a compliance signal.

Step three: Build the AI layer before you hire anyone

For every column-one task, write and save a prompt that produces a usable draft. Store them somewhere your future VA can access. This is the step agents skip, and skipping it is why they end up paying a person to do free work.

Step four: Let AI write the SOPs

Record yourself doing a column-two task once, out loud, start to finish. Feed the transcript to Claude and ask it to produce a numbered standard operating procedure with the decision points, the required inputs, and the definition of done. Edit for thirty seconds. That’s your training document. You’ve now converted a task you’ve never been able to explain into something a person can run on day one.

Step five: Set a review cadence and hold it

Pick a fixed time — end of day, twice a week, whatever your volume supports — and review completed output, not activity. If you’re checking in more often than that, the SOP isn’t finished. Go back to step four.

How I use this in my own business

I close 70+ transactions a year in San Antonio on roughly five hours a week of active management, and the three-column split is why the math works.

When I take a listing in the Stone Oak area, column one fires first. I run the property specs through a saved Claude prompt that writes the MLS description in my voice with fair housing language handled, and I edit it for about thirty seconds. Same prompt library produces the social captions, the email to my sphere, and the neighbor letter — all from one input.

Column two picks it up from there. My assistant confirms every factual detail in the AI draft against the actual property data, loads the listing, schedules the photographer, opens the transaction file, and sets the deadline tracking in the CRM. She works from an SOP that I dictated once into my phone and had AI turn into a checklist. I haven’t re-explained that process in over a year.

Column three is short and it’s mine. The pricing conversation. The listing appointment. The contract. Negotiation. Every showing. Those don’t move, no matter how good the system gets. The automation layer around the MLS handles the distribution — I handle the decisions.

The result isn’t that I work less on the hard parts. It’s that I stopped spending my week on the parts that were never hard to begin with.

Common mistakes

Hiring the VA before building the AI layer. You’ll pay a person hourly to do work a saved prompt does in eight seconds, and you’ll both be frustrated within a month.

Delegating tasks you’ve never documented. If you can’t write it down, you can’t hand it off — to a person or a machine. The inability to explain the task is the actual bottleneck.

Letting a VA make outbound prospecting calls. This is the most common and most expensive mistake in the whole framework. In Texas, soliciting listings is licensed activity, and the liability lands on you, not on them.

Skipping the verification layer. AI drafts confidently and wrongly. If nobody checks square footage, tax figures, or school zones against source data before it publishes, you’ve automated your errors instead of your work.

Managing activity instead of output. Daily check-ins feel like leadership. They’re actually a symptom of an incomplete SOP, and they put you right back in the middle of the business you were trying to step out of.

Treating the split as permanent. Re-audit quarterly. Tasks migrate from column two to column one as tools improve, and your VA’s scope should keep moving up the value chain.

Frequently Asked Questions

Can AI replace a real estate virtual assistant?

No. AI replaces specific tasks, not the role. It drafts, formats, summarizes, and researches — but it can’t verify that a listing description matches the actual property, chase a lender for a missing document, or make a judgment call when a deadline slips. AI shrinks the VA’s task list and raises the value of what’s left.

What tasks should I give a real estate virtual assistant?

Give a VA transaction file management, deadline tracking, CRM hygiene, scheduling, vendor coordination, inbox triage, social posting, and verification of AI-generated drafts. The common thread is judgment, follow-through, or a human on the other end. Anything with a predictable input and a template output belongs to AI instead.

What can an unlicensed assistant legally do in Texas?

TREC allows unlicensed assistants to perform administrative tasks: inputting data or typing contracts as specifically directed by a license holder, bookkeeping, ordering supplies and scheduling maintenance, and arranging showing appointments for the license holder who will show the property. They can’t show property, host open houses, solicit listings, or review contracts. Confirm your own state’s rules with your broker. HousingWireHousingWire

How much does a real estate virtual assistant cost?

Cost varies widely by market, scope, and whether you hire domestically or offshore, so treat any single number with skepticism. The better anchor is your budget ceiling: NAR’s 2026 Member Profile puts total median business expenses at $9,530. Narrow scope is what makes a VA affordable — which is exactly what the three-column split produces. Mile High Title Guy

How do I train a VA to use AI in my business?

Give them your saved prompt library and the SOPs, not general AI training. Record yourself doing each task once, have AI convert the transcript into a numbered procedure, and specify which prompt runs at which step. Your VA doesn’t need to be an AI expert. They need to run your prompts and verify the output.

Should I set up AI workflows before or after hiring a VA?

Before, without exception. Building the AI layer first tells you exactly what’s left over — and what’s left over is the real job description. Agents who hire first end up paying a person to do work a prompt handles for free, then conclude that VAs don’t work. The sequence is the whole difference.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

The delegation framework in this post is also the foundation of her Social Handoff keynote, a three-hour working session on handing social media to an assistant without losing your voice.

Book Emily to speak at your next event:
Email: eterrell@yourcoach.com
Phone: (210) 400-9191
Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals/year while coaching agents nationwide.

Real estate agents must disclose AI use when the output could mislead a client — digitally altered listing images, AI-generated video, and automated valuations. Drafting help on emails and descriptions generally requires no disclosure. California’s AB 723 made image disclosure law on January 1, 2026. This guide gives you the three-tier test and the workflow.

Key Takeaways

• Disclosure is triggered by what the AI output represents, not by whether you used AI at all.
• California’s AB 723 has been in effect since January 1, 2026, and requires a disclosure statement plus access to the original unaltered image.
• Wisconsin follows in 2027, and New York has a pending bill covering video and immersive media.
• Buyers already assume you’re using AI, and they punish AI errors harder than human ones — so concealment buys you nothing and costs you everything if it surfaces.
• Blanket “I used AI” labels on every email are not required and are not the goal — precision is.

What is AI disclosure in real estate?

AI disclosure in real estate is the practice of telling consumers when artificial intelligence has changed what they see, or has produced an output they may rely on to make a decision. It applies to listing photos, listing video, virtual staging, automated valuations, and AI-assisted screening. It does not generally apply to using AI as a drafting or research assistant behind the scenes.

The distinction matters because the legal trigger is misrepresentation, not automation. A tool that helps you write faster isn’t a disclosure event. A tool that changes what a buyer believes about a property is.

Why this matters for real estate agents

Adoption is finished. The argument is over. Delta Media Group’s third annual Real Estate AI & Leadership Survey, covering more than 100 brokerage leaders at firms responsible for over two-thirds of U.S. transactions, found that 97% said their agents are using AI — up from 80% in 2024, with 82% of agents using it to write listing descriptions (Real Estate News, January 2026).

Consumers know. Cotality’s AI in Housing 2026 Report found that three in four buyers expect AI somewhere in the transaction, and 80% assume real estate agents specifically are using it (HousingWire, April 2026).

Here’s the part most agents miss: assumption is not acceptance. In the same report, U.S. trust in AI to help find a home fell from 30% in 2025 to 16% in 2026, while 68% of buyers said they want clear notification whenever AI generates a listing, price, or mortgage recommendation — and 61% of baby boomers said it should be a legal requirement (Cotality, “Trust, but verify,” April 2026).

“Buyers are not asking whether AI is involved — they assume it is.” — John Rogers, Chief Data and Analytics Officer, Cotality

Then there’s the finding that should decide your policy on its own. Cotality measured what it calls the trust cliff: 70% of buyers say a significant AI-generated listing error would reduce their trust in the platform, compared with 60% for the identical error made by a human agent — a ten-point penalty that holds across every generation surveyed (Cotality, “Trust, but verify,” April 2026). You don’t get the grace a human gets. Publish an AI mistake and it costs you more than the same mistake made by hand.

The compliance gap is measurable and it’s ugly. A Coraly study of just under 40,000 primary listing images from Zillow, Redfin, Realtor.com, and Homes.com found 10.8% showed evidence of digital alteration — sky replacement accounted for 69% of those — and more than 90% of altered images carried no visible disclosure in the image, caption, listing description, or adjacent text (HousingWire, June 2026).

That’s the number a plaintiff’s attorney reads. That’s also the number a state legislature reads.

The three tiers of AI use — and which ones you disclose

Most agents are trying to answer one question (“should I say I used AI?”) when there are actually three, and they have different answers. Sort every AI task into one of these tiers before you publish anything.

Tier 1: Assistance — no disclosure required

AI drafts your follow-up email. AI summarizes a market report. AI turns your voice memo into a listing description that you then verify line by line. AI builds your content calendar. This is the tier I break down step by step in how to train ChatGPT for real estate agents.

Nothing here changes what a client believes about a property. Nobody discloses spellcheck, and nobody discloses a transcription tool. You own the output the moment you publish it — which means you’re responsible for its accuracy under your state’s advertising rules regardless of what wrote the first draft.

Tier 2: Representation — disclose, and in some states it’s law

This is where the legal exposure lives. California’s Assembly Bill 723, effective January 1, 2026, requires brokers, salespersons, or those acting on their behalf to disclose when digitally altered images are used and to provide access to the original, unaltered versions — and it applies when software or AI adds, removes, or changes visible elements including furniture, appliances, flooring, landscaping, façades, floor plans, window views, or neighboring properties (HousingWire, June 2026).

California draws a workable line: lighting, cropping, sharpening, and color correction are permitted as long as they don’t change how the property actually looks (HousingWire, June 2026).

Other states are moving. Wisconsin’s 2025 Act 69, effective in 2027, requires disclosure when advertising is altered using technology in a way that creates a false or misleading impression, and its scope covers reels, animations, and generated video — not just photos. New York regulators have warned that AI-generated listing imagery can produce misleading representations, and a pending bill, S9584, would define “digital representations” to include images, video, and immersive media (HousingWire, June 2026).

Video deserves its own sentence. An AI-generated video that simulates a drone approach or a walkthrough built from still photographs presents movement that was never captured — the images may be real, but the experience is not (HousingWire, June 2026). If a buyer thinks they’re watching footage, that’s a disclosure event.

Tier 3: Judgment — disclose and verify with a human

Automated valuations, AI-assisted tenant or buyer screening, and any AI output presented as analysis rather than marketing. This tier carries the highest liability and the lowest consumer tolerance.

AI tools are research and communication tools, not valuation tools — a point I’ve covered in detail in why agents who don’t understand AVMs keep losing listing appointments. If an AI-generated number reaches a client, that client needs to know what produced it and what its limits are.

Fair housing sits on top of all three tiers. The Fair Housing Act applies to your advertising no matter who or what drafted it. AI is fluent, confident, and completely unaware that “perfect for young professionals” or “great family neighborhood” can read as steering on a protected class. Every AI-drafted description gets a fair housing read before it goes anywhere near the MLS.

What Texas agents need to know right now

Texas has no AB 723 equivalent. That is not the same as having no rules.

TREC’s advertising rules were built to cover the media agents actually use: an advertisement on social media complies only if the license holder links to a profile or separate page carrying the required information, and subsection (d) of Rule 535.155 sets out 20 specific examples of advertisements that may mislead the public — with TREC noting those examples are not the only ways an advertisement can be misleading. Chapter 1101.652(b)(23) of the Texas Occupations Code states plainly that a license holder’s advertising cannot be misleading (TREC).

Translation: an undisclosed AI-altered listing photo in Texas doesn’t need a dedicated AI statute to be a problem. It’s already a misleading advertisement. Read TREC’s guidance directly, and check your MLS rules separately — many MLS systems require “virtually staged” labels independent of state law.

This is general information, not legal advice. AI disclosure rules vary by state, MLS, and brokerage policy, and they are changing quickly. Confirm your obligations with your broker and a licensed attorney in your state before you change your workflow.

How I use this in my own business

I run 70+ transactions a year on systems, and AI touches most of my marketing workflow. Here’s exactly where the disclosure line falls in my own San Antonio practice.

My listing descriptions start as AI drafts built from my own showing notes and the actual property data. I rewrite, verify every claim against the seller’s disclosure and my own walkthrough, and run a fair housing pass. No disclosure — because nothing in that output changes what a buyer believes about the house.

My photo workflow is different. On a recent listing, that image carried a label directly on the photo. I keep the original and the altered version in the transaction file, every time, on every listing — even though Texas doesn’t currently require me to.

That’s not caution for its own sake. It’s a one-line habit that costs me nothing and removes an entire category of future problems.

The systems thinking here is the same as everything else I teach: build the workflow once, and compliance stops being a decision you make under pressure. It’s also the same trust math I break down in how to stop letting AI erase you — the agents who win are the ones both humans and machines can verify.

Common mistakes

Disclosing everything. Slapping “AI-assisted” on every email and caption isn’t compliance — it’s noise, and it trains clients to discount the label when it actually matters. Precision beats volume.

Assuming your state’s silence is permission. Most states haven’t passed an AI statute. Nearly all of them already prohibit misleading advertising, which is the same trap with a different name.

Burying the disclosure. A label that lives in a listing’s fine print, three clicks from the image, doesn’t do the job. Disclosure travels with the media it describes.

Skipping the fair housing reading. AI writes clean, confident copy that can carry protected-class implications you’d never write yourself. You’re liable for it either way.

Not keeping the originals. Coraly’s researchers called the compliance gap a workflow problem rather than a willingness problem — agents receive final JPEGs from photographers with no metadata and no record of what changed, and the originals sometimes get deleted after delivery (HousingWire, June 2026). Ask your photographer for originals in writing, before the shoot.

Treating an AI valuation as a CMA. AI produces text about numbers. It does not produce a defensible opinion of value anchored in local market data.

Frequently Asked Questions

Do real estate agents have to tell clients they used AI?

Not for general assistance. There’s no requirement to disclose that AI helped draft an email, summarize a report, or produce a first-pass listing description you then verified. Disclosure obligations attach to outputs that could mislead — altered images, simulated video, and automated valuations presented as analysis. The trigger is misrepresentation, not automation.

Is it illegal to use AI-generated listing photos?

Using them isn’t illegal. Using them without disclosure can be. In California, AB 723 has required a disclosure statement plus access to the original unaltered image since January 1, 2026, whenever software or AI adds, removes, or changes visible elements of a property. Basic lighting, cropping, and color correction that don’t change how the property looks are permitted without disclosure.

Do I have to disclose AI if I’m not in California?

Check three sources: your state statute, your MLS rules, and your brokerage policy. Most states have no AI-specific law yet, but nearly all prohibit misleading advertising — which covers an undisclosed altered photo regardless. Wisconsin’s rules take effect in 2027 and New York has pending legislation, so the answer where you practice is likely to change.

Does using ChatGPT to write a listing description require disclosure?

Generally no, provided you verify it. The description is your advertisement the moment you publish it, and you’re accountable for its accuracy under your state’s advertising rules no matter what drafted it. The real risk isn’t the tool — it’s publishing unverified claims about square footage, features, or conditions that the AI inferred rather than confirmed.

Can AI-generated listing descriptions violate fair housing laws?

Yes, and this is the most underestimated risk in the whole category. AI produces confident, polished language that can imply preference for a protected class — phrases like “perfect for young professionals” or “great family neighborhood” can read as steering. The Fair Housing Act applies to your advertising regardless of who or what wrote it. Every AI draft needs a fair housing review.

What should an AI disclosure actually say?

Plain language, placed with the media it describes. “Virtually staged. Furniture shown is not included.” “Image has been digitally altered. Original photo available at the link provided.” “Video created from listing photographs. No drone footage was captured.” Legal-department phrasing isn’t required and usually performs worse. The test is whether a buyer immediately understands what was real and what was generated.

Does AI disclosure hurt client trust?

Some research suggests labeling AI-generated marketing content can reduce perceived authenticity, and Cotality found buyers penalize AI errors about ten points harder than identical human errors. That’s why blanket disclosure is the wrong strategy. Disclose where a client could rely on the output to their detriment, and stay quiet about backend drafting assistance. Precision protects both your compliance position and your credibility.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event: Email: eterrell@yourcoach.com Phone: (210) 400-9191 Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals/year while coaching agents nationwide.

AI-generated content can put your real estate license at risk, but not because you used AI — because you published something inaccurate, misleading, or discriminatory. Regulators in California, New York, and Texas hold the licensee and the broker responsible for the output, not the tool. Here’s what actually triggers discipline, state by state.

Key Takeaways

• No state disciplines a licensee for using AI. Every enforcement path runs through what the content says, not what wrote it.
• Texas agents are being told TRAIGA created an AI disclosure duty for them. It didn’t — the consumer notice requirement landed on government entities, not private licensees.
• California is the outlier with a real, specific rule: digitally altered listing images require disclosure and access to the original.
• The violation actually happening at volume is boring — AI drafts strip the broker’s name out of every caption, and TREC Rule 535.155 requires it on every advertisement.
• Your broker’s supervision duty extends to your AI tools. That makes an AI policy a brokerage-level problem, not an agent-level preference.

What is AI real estate compliance?

AI real estate compliance is the practice of reviewing, verifying, and documenting AI-generated marketing, communication, and transaction content before it reaches a consumer, so the output meets the same advertising, fair housing, and disclosure standards as anything you wrote yourself. It is not a separate body of law. It is an existing license law applied to a faster production process.

That distinction matters more than it sounds. Agents keep looking for the “AI rule” that tells them what they’re allowed to do. In most states, it doesn’t exist — and looking for it is why they miss the rules that already apply.

Why this matters for real estate agents

Most agents are producing at a volume their review process was never built for. According to NAR’s 2025 Technology Survey, released September 18, 2025, AI adoption reached 68% of agents, yet only 17% reported a significant positive impact on their business and 46% noticed no difference at all (NAR). Two-thirds of the industry is generating more content than ever and most of it isn’t moving anything.

Here’s the thing nobody wants to tell you: the same gap that makes AI unproductive is what makes it a license problem. Output you don’t review doesn’t convert, and output you don’t review is exactly what a regulator finds.

The California Department of Real Estate said it plainly in a licensee advisory issued March 17, 2026: a broker’s supervisory obligation extends to the tools used to conduct licensed or unlicensed activities, including AI-powered software, and when an AI tool produces inaccurate information or misleading advertising, responsibility rests with the licensee and their responsible broker rather than the technology provider (California DRE). The advisory goes further and compares using AI to perform licensed activity to handing that work to an unlicensed assistant.

“The tool never signs the listing agreement. You do. Every regulator that has weighed in has landed in the same place — the license holder owns the output, and ‘the AI wrote it’ has never once worked as a defense.” — Emily Terrell, Tom Ferry Coach

Which AI laws actually apply to real estate agents?

This is where most of the advice circulating right now is wrong. Four jurisdictions matter, and only two of them create a duty most agents don’t already have.

Does the Texas AI law require agents to disclose AI use?

No, and this is the most persistent myth in Texas right now. The Texas Responsible Artificial Intelligence Governance Act, HB 149, took effect January 1, 2026. Earlier drafts would have imposed a broad consumer AI-notice duty on private businesses. The enacted version narrowed that duty to governmental entities, with a separate obligation for healthcare providers in treatment contexts (Texas Legislature bill analysis). Texas real estate licensees are not covered by that notice requirement.

What still applies to you in Texas is the rule that was always there: TREC Rule 535.155. More on that below, because it’s the one that bites.

What does California require for AI-altered listing photos?

California created the clearest and most specific obligation in the country. As of January 1, 2026, Business and Professions Code Section 10140.8 requires licensees to include a disclosure when advertising images have been digitally modified in a way that changes the appearance of the property — including modifications created or enhanced by AI — and to make the original, unaltered image available to consumers (California Legislative Information).

The DRE advisory adds the part agents skip: failure to review AI-generated images for compliance can produce a violation even when the alteration was unintentional. You don’t get credit for not noticing.

Has New York taken a position on AI listings?

Yes. On November 13, 2025, the New York Department of State issued a trend alert warning that automated tools may produce misleading or exaggerated representations of properties, and that dishonest or misleading advertisements can subject a broker or salesperson to disciplinary action under Real Property Law Section 441-c (NY Department of State). No new statute. Existing deceptive advertising law, pointed directly at AI imagery.

What about Utah’s AI Policy Act?

Utah’s Artificial Intelligence Policy Act does reach “regulated occupations,” which is why agents keep hearing it applies to them. The 2025 amendments effective May 7, 2025 narrowed proactive disclosure to high-risk interactions built around health, financial, and biometric data, and advice in healthcare, law, and finance. Residential brokerage does not sit cleanly inside that definition. Utah licensees should confirm current scope with their broker and counsel rather than assuming either direction.

The violation nobody is writing about

Strip away the AI-specific headlines and the most common exposure for Texas agents right now has nothing to do with photos, fair housing, or state AI statutes. It’s the broker name.

TREC defines an advertisement as any form of communication by or on behalf of a license holder designed to attract the public to use real estate brokerage services — explicitly including email, text messages, social media, and the internet. Rule 535.155 requires every advertisement to include the name of the license holder or team placing it, plus the broker’s name in at least half the size of the largest contact information for any sales agent, associated broker, or team name in the ad (TREC).

Now think about what your AI tool hands you. Ask any model for thirty Instagram captions and you get thirty captions. Not one of them includes your broker’s name. The tool has no idea it’s required. If you’re batching a month of content and publishing straight from the output, you may have just produced thirty advertisements that don’t comply — and the failure has nothing to do with AI being dangerous and everything to do with nobody reviewing the draft.

That’s the actual risk profile. Not a robot writing something scandalous. A system with no review step, running at volume.

How I use this in my own business

I close 70+ transactions a year in San Antonio on roughly five hours of active management per week, and AI touches nearly every piece of marketing I produce. What makes that work isn’t restraint — it’s that the compliance step is built into the system instead of bolted on afterward.

Three things sit in front of every AI output before it publishes:

The brand and compliance block lives in the prompt, not in my memory. Every content prompt I use ends with the broker identification requirement and a fair housing constraint. That means the draft comes back closer to being compliant instead of needing a rewrite.

Every property fact gets verified against the MLS input, not against the draft. AI will confidently describe a third bedroom that doesn’t exist. I check square footage, bed and bath count, year built, and any feature claim against source data before anything goes out.

Nothing published from the tool. Output goes to a review queue, gets read once by a human, and gets released. That single step costs about ninety seconds per asset and eliminates most of what would otherwise be a violation.

That’s the whole system. It isn’t sophisticated. It’s just consistent, which is why it holds up at 70 transactions instead of collapsing at ten.

Common mistakes

Treating a disclaimer as a fix. A small line of text at the bottom of a listing does not correct an image that gives buyers a fundamentally wrong impression of the property. California requires the disclosure to accompany the altered image and requires access to the original.

Assuming your state has an AI rule because some state does. Four states have moved. Most haven’t. Acting on the wrong state’s requirement wastes effort and still leaves you exposed to the rule that actually applies.

Publishing AI captions straight from the tool. This is the broker-name problem, and it scales with your posting volume. The more disciplined your content calendar, the more advertisements you’re producing.

Letting AI generate copy about neighborhoods and buyers. Models produce phrasing like “perfect for young families” or “safe neighborhood” without hesitation. That language creates fair housing exposure regardless of who typed it, and the DRE advisory is explicit that AI use does not shield a licensee from liability under fair housing law.

Running an AI stack with no written brokerage policy. If a broker’s supervision duty covers the tools, then “we don’t have a policy” is itself the finding. The DRE lists written policies, training, and documentation of review steps among its recommended practices.

Frequently Asked Questions

Do I have to disclose that I used AI to write a listing description?

In most states, no. There is currently no general requirement that a licensee disclose AI authorship of listing text. What is required is that the description be accurate and not misleading. The obligation attaches to the content, not the authorship. Check your state commission and MLS rules, which can impose additional requirements beyond state statute.

Does the Texas AI law require real estate agents to disclose AI use?

No. TRAIGA, House Bill 149, took effect January 1, 2026, but the enacted version applied its consumer AI-notice duty to governmental entities rather than private businesses, with a separate provision for healthcare providers. Texas real estate licensees are not subject to that notice requirement. TREC’s existing advertising and misrepresentation rules still apply in full.

Can I get in trouble for using AI-enhanced listing photos?

In California, yes, if you don’t disclose. Business and Professions Code Section 10140.8, effective January 1, 2026, requires a disclosure when an image has been digitally modified in a way that changes how the property appears, plus access to the unaltered original. New York has warned that misleading AI imagery can trigger discipline under existing deceptive advertising law. Basic lighting and color correction generally aren’t the concern.

Is my broker responsible if AI generates a bad listing description?

Yes, in addition to you. The California DRE stated in March 2026 that a broker’s supervisory obligation extends to AI-powered software, and that responsibility for inaccurate or misleading AI output rests with the licensee and the responsible broker rather than the vendor. Brokers can face separate discipline for failing to reasonably supervise. This is why AI policy belongs at the brokerage level.

Does AI-generated content violate fair housing rules?

Not inherently, but it can produce violations quickly. AI models generate phrases signaling preference based on protected characteristics without any awareness that they’re doing it. Fair housing law applies to the advertisement regardless of who or what drafted it, and using an AI tool provides no defense. Build the constraint into your prompt and review every line before publishing.

What should a brokerage AI policy include?

At minimum: which tools are approved, what client data may never be entered into them, a required human review step before any AI output reaches a consumer, the advertising identification requirements for your state, a fair housing review standard, and documentation showing that reviews happened. The DRE specifically cites written policies, training, and documented compliance steps as evidence of good-faith effort.

Do I need to worry about entering client information into AI tools?

Yes. Many AI platforms retain and process what you input. Before using a tool, know what data it collects, how long it retains it, whether it’s shared with third parties, and what protections exist. Avoid entering confidential client financial details, transaction specifics, or personal identifying information into public or unsecured platforms.

This is general information, not legal advice. AI and advertising rules vary by state and change frequently. Consult your broker, your state real estate commission, and an attorney licensed in your state before setting policy or acting on anything in this post.

Related reading

• Real Estate MLS Automation: What Agents Can Automate
• How to Automate MLS Listing Syndication in 2026
• Real Estate Video Editing: The Batch System That Works

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event: Email: eterrell@yourcoach.com Phone: (210) 400-9191 Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Speaker for NAHREP, eXp Con, and brokerages nationwide.

AI training for real estate agents should cover four things: one live build on the agent’s own listing, prompt architecture instead of a prompt list, compliance guardrails around Fair Housing and TREC advertising rules, and a handoff into the CRM. Tool tours don’t change behavior. This guide gives the full agenda, timing, and pre-work.

Key Takeaways

• Adoption isn’t the problem anymore — 68% of agents already use AI, and most report it hasn’t changed their business.
• One live build on real material beats five tool demos, because watching produces admiration and building produces adoption.
• Teach prompt architecture, not a prompt list — the list expires with the next model release.
• Compliance is a required segment, not a closing disclaimer, and it’s the part brokerage leaders actually buy.
• A session that ends without one named workflow and a check-in date ends with nothing.

What is AI training for real estate agents?

AI training for real estate agents is a working session that installs one repeatable AI workflow into an agent’s actual business, using their own listings, leads, and CRM. It is not a software demonstration and it is not a keynote about the future of technology. The measure of a good session is what’s still running 30 days later, not how the room felt when it ended.

That distinction matters more than it sounds. A room that’s buzzing at the end usually means agents watched something impressive. The rooms that produce change are the ones where a third of the audience got visibly frustrated fighting with their own prompt — and then got it working.

Why this matters for real estate agents

The gap between using AI and getting anything from it is now the entire story. According to NAR’s 2025 Technology Survey (September 2025), only 17% of agents reported AI had a significantly positive impact on their business, while 46% said it had no noticeable impact at all. That same survey found 20% of agents use AI tools daily and 22% use them weekly.

Read those two numbers together. Nearly half the agents touching AI every week are getting nothing measurable out of it. That’s not a technology problem. That’s a training problem.

The economics make it urgent. According to NAR’s 2026 Member Profile (June 2026), the typical agent closed nine transaction sides in 2025 with a median gross income of $59,200, while median business expenses climbed to $9,530 from $8,010 the year before. Expenses are rising faster than income. Hours spent drafting listing copy and follow-up emails are hours that don’t get billed and don’t get recovered.

“An AI output that doesn’t land in your CRM is a party trick. The workflow only counts if it survives the drive home.” — Emily Terrell, Tom Ferry Coach

What should an AI training session for agents actually cover?

What should the first ten minutes do?

The first ten minutes establish the cost, not the capability. Before anyone opens a tool, every agent in the room writes down roughly how many hours went into listing copy, follow-up drafting, and social captions last month. Their number, on paper.

That number does two jobs. It’s the hook — agents underestimate it badly until they add it up. And it’s the baseline you measure against at the end, which is the only honest way to report the session’s value back to the brokerage that paid for it.

Then one rule, stated once and never revisited: AI drafts, the agent decides. Everything downstream inherits that frame, including the compliance segment.

Should the session teach prompts or prompt architecture?

Architecture, always. A prompt list has a shelf life of about one model release. The anatomy underneath it doesn’t expire: role, context, constraints, output format, example.

The way to teach it is a teardown, not a lecture. Put one weak prompt and one strong prompt side by side on the same task, run both, and let the room see the difference in the output. Then name what changed. Agents take a five-part list on faith; they take a visible quality delta as fact.

This is also what separates real training from the “10 ChatGPT prompts for Realtors” circuit. One hands out fish. The other teaches the pattern that generates the next hundred prompts. For a deeper walkthrough on customizing a tool to your voice and files, see how to train ChatGPT for real estate agents.

Why does the live build have to use the agent’s own listing?

Because hypothetical material produces hypothetical adoption. When the room builds against a made-up three-bedroom in a made-up subdivision, everyone succeeds and nobody learns anything transferable. When they build against a listing they’re actually sitting on, they hit the real friction — missing data, a detail they can’t verify, a tone that doesn’t sound like them — and solving that friction in the room is the entire point.

Budget 20 to 25 minutes for the build. That feels absurd on a 60-minute agenda until you accept that it’s the only segment that changes behavior.

Run hard checkpoints inside it — roughly every seven minutes, the room stops and each agent reads their current output to the person next to them. Pair rescue, not hand-raising. Agents won’t raise a hand in a room of peers to admit they’re lost. They’ll tell one person sitting beside them.

What compliance topics must an AI training cover?

Four, as a full segment with a worked example — not a slide of warnings at the end:

• Fabricated property facts. AI will confidently invent a square footage, a school district, or a roof age. Every generated fact gets verified against the MLS record before it is published.
• Fair Housing exposure. Generated listing copy can drift into language signaling preference based on protected class. Review the NAR Fair Housing resources and build the constraint into the prompt itself rather than catching it in edit.
• Texas advertising and disclosure rules. TREC Rule 535.155 governs how Texas license holders advertise, and AI-generated marketing is still advertising. Agents outside Texas need the equivalent rule from their own commission.
• Client data in consumer AI tools. Pasting a client’s financial details or contact record into a general-purpose chatbot is a decision with consequences. Cover what goes in and what never does.

This is general information, not legal advice. Consult your broker or attorney on specific listings and campaigns.

Brokerage leaders and event planners buy this segment more than any other, because it’s their liability sitting in the room. Most AI trainers skip it entirely.

Where does the AI output go after the session?

Into a system, or it evaporates by Thursday. The last working block of the session wires the output somewhere it persists — a Follow Up Boss template, a saved prompt library, an email sequence. Faster content production without a system underneath it is just faster chaos.

If the workflow being built is follow-up, that means it lands in the CRM before anyone leaves. Which real estate CRM works best with AI covers the integration questions worth answering before the session, not during it.

The 60-minute agenda

Minutes	Segment	Outcome
0–5	Their number, on paper	Baseline hours captured
5–10	One rule: AI drafts, you decide	Frame set
10–18	Prompt anatomy via side-by-side teardown	Pattern named
18–40	Live build on their own listing, three pair checkpoints	Working output
40–48	Guardrails with a worked example	Risk understood
48–55	Wire the output into a system	Destination confirmed
55–60	One workflow, one owner, one date	Commitment made

At 45 minutes, cut the CRM block to a single named next step and compress prompt anatomy to the teardown alone. Keep the build and keep compliance — the build is what gets you rebooked by agents, compliance is what gets you rebooked by the brokerage.

At half a day, this agenda becomes module one. Add a second build on a different task, full prompt-library construction, and a block where agents configure the workflow in their CRM while you’re standing there.

What the pre-work has to require

Sent 72 hours out and enforced by the host, not the speaker:

• A real active listing or a real cold lead, open on their device
• A working account on the tool being used, logged in and tested before arrival
• Last month’s rough hours on listing copy, follow-up, and captions

Without this you don’t have training. You have a demo. Build a five-minute buffer before the session for the agents who ignored the email, and pair them with someone who didn’t rather than stopping the room.

Three tech failures kill live builds reliably: venue wifi collapsing under 80 simultaneous connections, someone hitting a paywall mid-build, and mobile-only attendees who can’t work in two windows. Have a fallback that lets a stalled agent follow a partner’s screen without losing the thread.

How I use this in my own business

I close 70+ transactions a year in San Antonio in roughly five hours a week of active management, and the reason that math works is that every AI workflow I use has a destination. My listing marketing suite doesn’t live in a chat window — it moves into the templates my transaction coordinator already uses, which is why it survives a week where I’m on a stage instead of at my desk.

That’s the same reason I stopped running tool tours from the stage. Agents would come up afterward genuinely energized, and 30 days later nothing had changed in their business. Now the room builds against their own material, or I don’t run the segment.

The pre-work requirement costs me registrations at some events. Organizers push back on asking agents to prepare anything. I’d rather have 60 agents who arrive ready than 200 who watch. The related principle for organizers is covered in building your real estate event like a system, not a show.

Common mistakes

1. Covering five tools instead of installing one workflow. Coverage is not the constraint. Retention is. Agents implement one thing, maybe two.
2. Demoing instead of building. The moment the room is watching you, you’ve traded behavioral change for applause.
3. Treating compliance as a closing disclaimer. It’s a segment with a worked example, or agents will assume it doesn’t apply to them.
4. Handing out a prompt list. It’s obsolete within months and it teaches nothing transferable.
5. Ending with energy instead of a date. No named workflow, no owner, no check-in — nothing survives the drive home.
6. Skipping the CRM handoff. Content production without a system is faster chaos, not leverage. The same principle applies to automating MLS listing syndication.

Frequently Asked Questions

How long should an AI training session for real estate agents be?

Sixty minutes is the working minimum for a session that includes a live build. Anything shorter forces you to demo instead of build, which produces interest but not adoption. Half-day formats allow two builds plus CRM configuration. A 45-minute keynote slot can work if you cut the CRM block and keep the build and compliance segments intact.

What should agents bring to an AI training?

A real active listing or a real cold lead open on their device, a logged-in and tested account on the tool being used, and their rough hours from last month on listing copy, follow-up, and captions. Hypothetical material produces hypothetical adoption. Without real material in front of them, agents watch rather than build.

Should AI training cover ChatGPT, Claude, or another tool?

Pick one tool for the live build and teach architecture that transfers. NAR’s 2025 Technology Survey found ChatGPT is used by 58% of agents who use AI, followed by Gemini at 20%. Building in the tool most of the room already has removed a setup barrier. The prompt structure you teach works across all of them.

How do you measure whether AI training worked?

Measure the number of agents running a named workflow 30 days later, not room energy or survey scores. Capture baseline hours at the start of the session so you have a before number. Send the host a recap within 48 hours listing attendance, baseline hours, workflows committed, and guardrails covered.

Does AI training need to cover compliance?

Yes, as a required segment. Generated listing copy can carry Fair Housing exposure, and AI-generated marketing is still advertising under state rules like TREC Rule 535.155 in Texas. Fabricated property facts and client data entered into consumer AI tools carry their own risk. This is general information, not legal advice — consult your broker or attorney.

What’s the difference between an AI keynote and an AI training?

A keynote changes what agents believe. Training changes what agents do. The structural difference is whether the audience builds something on their own material during the session. If nobody opens a laptop, it’s a keynote — which has its place, but shouldn’t be booked when the goal is adoption.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event: Email: eterrell@yourcoach.com Phone: (210) 400-9191 Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Licensed since 2016. Closing 70+ deals/year while coaching agents nationwide.

You can monetize a real estate YouTube channel two ways: YouTube’s ad program, which requires 1,000 subscribers and 4,000 watch hours, or client transactions, which requires neither. For licensed agents, the second path pays more and carries less regulatory risk. This guide runs the math and flags the RESPA and TREC problems buried in affiliate income.

Key Takeaways

• YouTube’s ad revenue program requires 1,000 subscribers plus 4,000 valid public watch hours in the trailing 12 months, or 1,000 subscribers plus 10 million Shorts views in 90 days.
• Watch time from Shorts and from unlisted or private videos does not count toward the 4,000-hour threshold — which disqualifies most agent listing libraries.
• Affiliate and sponsorship money from lenders, title companies, and inspectors is the single most dangerous revenue idea in this category for a licensed agent.
• The typical Realtor completes 10 transaction sides a year. One additional side is worth more than a year of ad revenue on a channel of that size.
• Build the channel as a lead system first. Ad revenue, if it arrives, is a rounding error you accept — not a strategy you chase.

What does it mean to monetize a real estate YouTube channel?

Monetizing a real estate YouTube channel means converting views into revenue, and there are three distinct mechanisms: YouTube Partner Program ad revenue and fan funding, third-party affiliate or sponsorship payments, and client transactions generated by viewers who hire you. Only the third is a real estate business model. The first two are creator business models that happen to be running on a real estate channel.

That distinction matters more than it sounds. It determines what content you make, what metrics you watch, and — for a licensed agent — which federal and state regulations you just walked into.

Why this matters for real estate agents

Here’s the thing nobody wants to tell you: most agents asking how to monetize a YouTube channel are optimizing the wrong number.

According to NAR’s 2025 Member Profile (August 2025), the typical Realtor completed 10 transaction sides in 2024 with median sales volume of $2.5 million. According to that same profile, the median gross income for Realtors rose to $58,100 in 2024 from $55,800 in 2023. Divide it out. The typical agent’s gross income per transaction side sits near $5,800.

Now look at what it takes to unlock ad revenue at all: 1,000 subscribers and 4,000 watch hours. That’s a year or more of consistent production for most local agents. At the end of it, you get a share of ad revenue on a channel whose audience is measured in the low thousands.

One extra closing beats it. Not marginally — decisively. And the content that produces that closing looks different from the content that produces watch hours.

“If your YouTube channel produced one additional transaction side this year, it outperformed the ad revenue from ten thousand views. Build for the closing, not the CPM.” — Emily Terrell, Tom Ferry Coach

The three revenue paths, and what each one actually costs

What does YouTube’s ad program actually require?

Full ad revenue access through the YouTube Partner Program requires 1,000 subscribers with either 4,000 valid public watch hours in the last 12 months, or 10 million valid public Shorts views in the last 90 days. You also need no active Community Guidelines strikes, two-step verification enabled, advanced features access, and a linked AdSense for YouTube account. YouTube says review typically takes about a month.

Three details disqualify a lot of agent channels immediately.

Watch hours from Shorts in the Shorts Feed do not count toward the 4,000-hour threshold. Watch hours from private, unlisted, or deleted videos do not count either — and unlisted is exactly how most agents publish listing tours for their sellers. And if a channel goes six months or more without an upload or a post, YouTube reserves the right to remove monetization.

YouTube also runs an expanded early-access tier with lower thresholds for fan funding and Shopping features in eligible countries. Check the Earn section of your own YouTube Studio for your current eligibility rather than trusting a third-party number.

What does affiliate and sponsorship income cost you in compliance risk?

This is the path most “monetize your channel” advice recommends, and for a licensed real estate agent it is the one that can cost you a license.

The standard pitch: put affiliate links to your preferred lender, title company, home warranty provider, or inspector in the video description and collect a commission when viewers use them. Read that sentence again with RESPA in front of you.

Under Regulation X, 12 CFR 1024.14, no person may give or accept any fee, kickback, or thing of value pursuant to an agreement or understanding that business incident to or part of a settlement service involving a federally related mortgage loan will be referred to any person. Lending, title, appraisal, and inspection are settlement services. A commission paid to you because a viewer used your link is compensation connected to a referral.

Texas layers on its own requirement. Under TREC Rule 535.155, an advertisement includes electronic media, social media, and the internet, and if you offer, recommend, or promote a service provider and expect compensation when a party uses that service, the advertisement must disclose that you may receive the compensation. The same rule requires each advertisement to carry the name of the license holder placing it and the broker’s name in at least half the size of the largest agent contact information in the ad. Your YouTube channel banner and video descriptions are advertisements.

And the FTC’s Endorsement Guides at 16 CFR Part 255 require any material connection — affiliate commissions included — to be disclosed clearly and conspicuously, in a way that’s difficult to miss and unavoidable in the communication itself. Not in a link. Not below the fold.

Non-settlement sponsorships — a camera brand, a CRM, an AI tool — sit on much safer ground, still require FTC disclosure, and still require your broker’s sign-off.

This is general information, not legal advice. Before you accept any compensation tied to a service provider, run the arrangement past your broker and a real estate attorney.

What does the transaction path pay?

The channel monetizes when a viewer becomes a client. That’s it. And it’s the only path where the asset compounds: a neighborhood video published eighteen months ago still gets found, still gets watched, and still produces a call.

The build is straightforward. Make videos that answer the questions a mover types into search — not the questions an audience finds entertaining. Put a single capture mechanism in every video: a specific address for a local guide, a booking link, one phone number. Route every inquiry into your CRM the same day with a named owner and a follow-up sequence. Measure inbound conversations, not subscribers.

I’ve written the full discoverability system in YouTube Is Not a Social Media Channel — It Is a Search Authority System, and the ranking mechanics in The Real Estate Agent’s Definitive Guide to YouTube Video SEO.

How I use this in my own business

My channel has never paid me a dollar in ad revenue, and I’ve never tried to make it.

What it does is answer questions for people relocating to San Antonio before they ever call an agent. A single video on what it’s actually like to live in Stone Oak — commute times, HOA realities, school boundaries, what the property tax bill looks like at that price point — has produced more buyer conversations than any paid lead source I’ve tested. Every one of those conversations lands in Follow Up Boss the same day it comes in, tagged by source, with a follow-up sequence already assigned.

I batch the filming. I don’t chase the algorithm. And I don’t measure the channel by subscriber count, because subscriber count has never once shown up on a settlement statement. If you want the production system behind that, it’s in Real Estate Video Editing: The Batch System That Works.

Common mistakes

Publishing listing tours as unlisted and wondering why watch hours stall. Unlisted watch time doesn’t count toward YPP thresholds and unlisted videos don’t get discovered. If the seller consents, publish publicly.

Taking a lender or title affiliate deal without running it past your broker. The revenue is small. The exposure is not.

Chasing Shorts views for monetization. Shorts-feed watch time doesn’t count toward the 4,000-hour threshold, and the Shorts path to full ad revenue requires 10 million views in 90 days.

Building for an audience instead of a market. A national audience of agents watching your content generates zero commission. A hundred local movers generate a pipeline.

Publishing without a capture mechanism. A video with no single, specific next step is a brand impression, not a lead source.

Letting the channel go dark for six months. Beyond losing the algorithm, YouTube reserves the right to remove monetization from channels inactive for six months or more.

Frequently Asked Questions

How many subscribers do you need to monetize a real estate YouTube channel?

Full ad revenue through the YouTube Partner Program requires 1,000 subscribers combined with either 4,000 valid public watch hours in the past 12 months or 10 million valid public Shorts views in the past 90 days. YouTube also runs an expanded early-access tier with lower thresholds for fan funding features in eligible countries. Check your current eligibility in the Earn section of YouTube Studio.

Is YouTube ad revenue worth it for a real estate agent?

Rarely, on its own. According to NAR’s 2025 Member Profile, the median gross income for Realtors was $58,100 across a typical 10 transaction sides in 2024 — roughly $5,800 per side. A local agent channel large enough to qualify for ad revenue will not produce that from ads. Treat ad revenue as a bonus, not a business case.

Can real estate agents use affiliate links on YouTube?

Not without serious care. RESPA Section 8 prohibits accepting a fee or thing of value for referring settlement service business, which covers lenders, title companies, appraisers, and inspectors. Non-settlement products like software or equipment are safer but still require FTC material-connection disclosure and your broker’s approval. Consult your broker and an attorney before accepting any of it.

Do TREC advertising rules apply to a YouTube channel?

Yes. TREC Rule 535.155 defines an advertisement to include electronic media, social media, and the internet, so channel branding, video descriptions, and listing content promoting brokerage services qualify. Each advertisement must include the name of the license holder placing it and the broker’s name at no less than half the size of the largest agent contact information shown.

Do YouTube Shorts count toward monetization for real estate agents?

Partially, and not the way most agents assume. Watch hours from Shorts in the Shorts Feed do not count toward the 4,000 valid public watch hours threshold. Shorts qualify through a separate path requiring 10 million valid public views in 90 days alongside 1,000 subscribers, which is out of reach for nearly every local real estate channel.

How long does it take to monetize a real estate YouTube channel?

For ad revenue, plan on twelve to eighteen months of consistent public long-form publishing to reach 1,000 subscribers and 4,000 watch hours, then roughly a month for YouTube’s review. For transaction revenue, the timeline is far shorter — a well-targeted local search video can produce an inbound conversation within weeks of publishing.

What should a real estate agent measure instead of subscribers?

Track inbound conversations sourced to video, appointments set from those conversations, and closings attributed to the channel. Log every one in your CRM with a source tag. Subscriber count and view count tell you the algorithm found you; only conversation count tells you the channel is producing business you can bank.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event: Email: eterrell@yourcoach.com Phone: (210) 400-9191 Web: coachemilyterrell.com

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.

By Emily Terrell — Top Coach and Speaker at Tom Ferry International. Real estate’s leading voice on AI, systems, and social media.

The best TikTok content ideas for real estate agents are hyperlocal price comparisons, reverse tours that withhold the number, client objections answered verbatim, transaction process breakdowns, and near-miss deal stories. Each one earns watch-through because the viewer already has the question. This guide covers the five formats, the weekly filming system, and the TREC compliance layer.

Key Takeaways

• Only 16% of Realtors use TikTok professionally, while 37% of U.S. adults and 63% of adults under 30 are on the platform — the widest adoption gap in real estate social media.
• Ideas aren’t the constraint. The same idea dies or travels depending on whether the first two seconds carry information or throat-clearing.
• Five repeatable formats cover a full year of content: price contrast, reverse tour, objection answer, process breakdown, and near-miss story.
• TikTok produces attention, not signed clients. The capture system lives off-platform, and if you don’t build it, you’ll quit at month four with good numbers and no closings.
• Automated DMs triggered by comment keywords are almost certainly advertisements under TREC Rule 535.155 and need broker identification inside the message.

What are TikTok content ideas for real estate agents?

TikTok content ideas for real estate agents are repeatable short-form video formats built around questions local buyers and sellers already have — pricing, neighborhoods, process, and the parts of a transaction nobody explains publicly. The strongest ideas aren’t creative. They’re structural: a format you can refill with new material every week without inventing anything.

That distinction matters more than agents expect. A format is scalable and repeatable. An idea is a one-off you have to replace next Tuesday.

Why does TikTok matter for real estate agents right now?

Because the gap between where agents are posting and where consumers are watching is larger on TikTok than anywhere else.

According to the National Association of Realtors’ 2026 Member Profile, the top platforms agents use professionally are Facebook (76%), Instagram (57%), LinkedIn (55%), YouTube (31%), and TikTok (16%). Meanwhile, Pew Research Center reports that 37% of U.S. adults use TikTok, rising to 63% of adults under 30.

Sixteen percent of agents. Sixty-three percent of the under-30 audience. That’s not a trend argument — that’s an arithmetic argument.

NAR’s own coverage makes the same point, noting that many agents are still leaving TikTok out of their marketing strategy even as younger buyers increasingly research neighborhoods, local businesses, and agents themselves on short-form video before they ever make contact.

Here’s the thing nobody wants to tell you: the reason most agents skip TikTok isn’t the algorithm. It’s that they think they need to be entertaining. They don’t. They need to be specific.

“Agents don’t lose on TikTok because their information is bad. They lose because the first two seconds of the video are spent being polite. Say the number, name the street, and let the value carry the video.” — Emily Terrell, Tom Ferry Coach

What are the five TikTok content formats that work for agents?

Rotate these five. Each one refills indefinitely from your actual business.

What does a hyperlocal price comparison video look like?

Two neighborhoods, one budget, side by side. “This is what $450,000 buys in Stone Oak. This is what it buys in Alamo Heights.” Open with the number on screen and in your first spoken line.

This format works because the viewer arrives already holding the question. You’re not creating curiosity — you’re answering demand that already exists. It also carries built-in search language: the neighborhood names are the keywords people type.

Refill it by rotating budget tiers and neighborhood pairs. One listing area gives you a dozen videos.

Why does the reverse tour outperform a standard listing walkthrough?

Because a standard tour answers the only interesting question in the first three words. “Just listed at $625,000” gives away the payoff before the viewer has committed anything.

Flip it. Show the house first. Show the detail people will react to. Deliver the price at second twenty. The unanswered question is what holds attention through the middle of the video, where most listing content collapses.

One rule: the price has to land. If you tease and then don’t deliver, you train people to leave early next time.

How do you turn client objections into TikTok videos?

Take the sentence a client actually said this week and answer it in forty seconds. “My lender told me I need twenty percent down.” “The listing agent said they already have an offer.” “Zillow says my house is worth more than that.”

One objection per video. Never two. The specificity is the hook — a real sentence sounds different from a manufactured one, and viewers can hear the difference.

Your CRM is the content calendar here. Every buyer and seller conversation you have this month is a script you already wrote.

What process breakdown content should agents post?

Explain the parts of a transaction that consumers never see. What an inspection actually finds. Why an appraisal came in low and what happens next. What the days between contract and close really look like. The difference between pre-qualification and pre-approval.

This is the most underused format in real estate short-form video, and it’s the one that builds the most authority. Nobody outside the business knows this material, and everybody buying or selling desperately wants it.

Does storytelling work on TikTok for real estate?

It does, in one specific shape: the near-miss. “This deal almost died on Thursday. Here’s what saved it.”

The story carries retention on its own architecture — the viewer stays for the resolution. Strip identifying details, get permission where the client is recognizable, and keep it to a single turning point.

Skip the lifestyle content. Nobody is following a real estate agent for aesthetics. They’re following you because you know something about where they want to live.

How do you film a month of TikTok in one afternoon?

Posting feels impossible when every video starts from a blank page. It stops feeling impossible when you batch.

Task	Time	Tool
Script fifteen hooks and CTAs	20 minutes	Claude or ChatGPT
Film fifteen videos, one outfit	90 minutes	Phone, natural light
Cut and caption	45 minutes	CapCut
Schedule the batch	15 minutes	Native scheduler
Daily comment replies	10 minutes/day	TikTok app

That’s roughly three and a half hours for a month of content at three posts per week. AI doesn’t create your personality — it removes the friction between having something to say and having it scripted. You still have to be on camera.

If you’re editing across platforms, the vertical master you cut for TikTok also serves Reels and Shorts. My batch editing system covers the export workflow so you’re not re-cutting the same footage three times.

How do you turn TikTok views into actual pipeline?

You don’t. Not on TikTok.

This is the part that gets agents in trouble, and it’s worth being blunt about: TikTok is a discovery engine, not a lead source. It produces reach and familiarity. It does not produce a signed buyer representation agreement. The conversion happens somewhere else, and if you haven’t built that somewhere else, the reach evaporates.

The capture layer has three parts:

One local asset worth trading an email for. A neighborhood guide, a school-boundary breakdown, a relocation checklist for your specific market. Not a generic buyer guide — those don’t convert because they’re available everywhere.

One route off the platform. Link in bio to a single landing page. Not your website homepage. Not six links. One.

One place the contact lands. In my business that’s Follow Up Boss with a dedicated smart list and an automated first-touch sequence. The tool matters less than the fact that a human follows up within twenty-four hours.

Views without this are a hobby. That’s the whole difference between an agent who posts for six months and quits and one who builds a pipeline out of it. You don’t need more leads — you need a better system for the ones you have, and that principle applies to the ones TikTok sends you too.

For the broader platform-selection logic, my post on social media strategy for new agents covers how to pick two platforms and ignore the rest until they’re working.

What TREC rules apply to real estate TikTok content?

Texas license holders should read TikTok as advertising, because TREC does.

Under TREC Rule 535.155, an advertisement is any communication by or on behalf of a license holder designed to attract the public to use real estate brokerage services — and the rule names social media, electronic media, and text messages explicitly. Each advertisement must include the name of the license holder placing it and the broker’s name in a readily noticeable location, at least half the size of the largest agent or team contact information in the ad.

Three practical consequences for TikTok:

Your profile is an ad. The broker name goes in the bio, not just on your website.

Comment-keyword automations are ads. If someone comments “guide” and an automated DM fires back with your offer, that DM is a communication designed to attract the public to brokerage services. Build broker identification into the message template itself, not into a follow-up.

A video about a specific listing is an ad. On-screen broker identification, every time.

The rule carves out communication to a license holder’s current client. A stranger who commented on a video is not a current client.

This is general information, not legal advice. Confirm your specific advertising setup with your broker or a Texas real estate attorney before you automate anything.

How I use this in my own business

I run three posts a week and I film them in one sitting, usually a Tuesday afternoon between appointments.

The format that has consistently outperformed everything else for me is the price contrast. When I was working at a Stone Oak listing earlier this year, I filmed a walkthrough of what that budget bought there and then drove twelve minutes and filmed the comparable at the same number in a different pocket of 78258. Same day, same shirt, two videos. The comparison video did several multiples of what the straight listing tour did, and three of the DMs that came out of it were relocation buyers who had never heard my name.

None of them closed off TikTok. They closed because they landed in Follow Up Boss and got a call. That’s the sequence, and I’d rather you see the whole sequence than just the video.

I closed 70+ transactions last year in roughly five hours a week of active management. Short-form video is one input into that, not the engine. The engine is the system underneath it.

Common mistakes

Opening with a logo animation. You’ve spent your entire hook window on branding. The face and the first sentence are the hook.

Posting listing tours exclusively. Your listings are interesting to about eleven people. Your market knowledge is interesting to everyone thinking about moving there.

Chasing trending audio. Trend audio buys you a marginal distribution bump and costs you the first three seconds of spoken information. Bad trade for a local business.

Two CTAs in one video. Follow me and comment and check the link in bio produces none of the three. Pick one.

Abandoning the format after one video performs. The correct response to a video that works is to make eleven more in the same shape. Most agents do the opposite and change topics.

Ignoring the broker identification requirement until someone complains. Fix it in the template once, not video by video.

Frequently Asked Questions

How often should real estate agents post on TikTok?

Three times a week is the working minimum, and consistency beats volume. Three well-structured videos weekly will outperform daily posting that runs out of steam by week three. Batch-film them in one session so posting frequency isn’t dependent on how you feel that morning. Consistency over seven days is what compounds.

Do real estate agents actually get clients from TikTok?

Agents get discovered on TikTok and convert clients elsewhere. The platform generates reach, familiarity, and inbound curiosity, particularly from relocation buyers and people under 40. The signed client comes from what happens after: a local resource worth an email address, one link off-platform, and a CRM with human follow-up inside twenty-four hours.

What should I post on TikTok if I have no listings?

Neighborhood and process content, which is stronger anyway. Price comparisons across areas, school-boundary explanations, what an inspection actually turns up, and answers to the questions buyers ask you daily. None of this requires inventory. New agents often outperform experienced ones on TikTok precisely because they build on knowledge rather than listings.

Is TikTok or Instagram better for real estate agents?

They serve different jobs. Instagram converts an audience that already knows you; TikTok reaches people who don’t. NAR’s 2026 Member Profile puts agent Instagram use at 57% versus 16% on TikTok, which means Instagram is more crowded with agents and TikTok has more open distribution. Run both from one vertical master file.

Do I have to show my face on TikTok?

You don’t have to start there. Voiceover paired with neighborhood footage, listing walkthroughs, or screen recordings works, and it’s a reasonable on-ramp. But familiarity is the asset short-form video builds, and familiarity requires a face and a voice eventually. Plan to be on camera by month two.

Do TikTok videos count as real estate advertising in Texas?

Yes. TREC Rule 535.155 defines an advertisement broadly enough to include social media and electronic communications designed to attract the public to brokerage services. Your profile, your listing videos, and your automated DMs all qualify. Broker identification is required in a readily noticeable location. Confirm your specific setup with your broker.

How long should real estate TikTok videos be?

Twenty to sixty seconds for teaching content, and shorter is usually safer. One tactic, one takeaway, no padding. If a topic needs more than sixty seconds, it’s two videos — and splitting it gives you a series people follow rather than one video they abandon at second thirty.

Bring this to your team or event

Emily Terrell speaks at brokerage events, real estate conferences, and team trainings on AI, systems, and social media — the exact playbook in this post, delivered live to your audience. As a Top Coach and Speaker at Tom Ferry International and an active agent closing 70+ transactions a year, Emily speaks from the stage about what’s working right now, not theory. Recent stages include NAHREP and eXp Con.

Book Emily to speak at your next event: Email: eterrell@yourcoach.com Phone: (210) 400-9191 Web: coachemilyterrell.com/keynote

For real estate agents who want to implement this: Get the weekly real estate prompt library at weeklyrealestateprompts.com or follow @coachemilyterrell on Instagram for daily systems and AI breakdowns.